Colorful earth model

Climate models help predict climate change trends

Caption: The paint-like swirls of this visualization from Los Alamos National Laboratory depict global water-surface temperatures, with the cooler temperatures designated by blues and warmer temperatures by reds, as delineated by recent climate modeling efforts. Credit: Los Alamos National Laboratory, FlickrCC.

Dear EarthTalk: How does climate modeling work? What is the state-of-the-art in the field these days, and what do these most recent models tell us about our future? – J.M., Austin, TX

Climate models are 3D figures of the Earth’s surface that demonstrate the cycling of energy and materials through the atmosphere, oceans and land. They compile geography, physics, chemistry and biology to analyze historical data and predict future global conditions. Ultimately, these models allow researchers and the public to explore Earth’s systems, climate change and the impacts of human activity on the planet.

The Coupled Model Intercomparison Projects (CMIP), conducted by the Intergovernmental Panel on Climate Change (IPCC), are some of the most detailed and expansive climate models to date. They show that the warming predicted from this century may be 0.4 degrees C greater than what was deduced from the CMIP5 in 2013. This may not sound significant, but it takes an immense amount of heat and, in this case, trapped greenhouse gases to warm the atmosphere, oceans and land that much. In the 1700s, it only took a 1-2 degree drop in global temperature to plunge the Earth into the Little Ice Age.

The CMIPs have also been proven to be very accurate. Fourteen of the 17 models done between 1970 and 2007 made similar projections of temperature change through the next decade as found by UC Berkeley researchers. These findings have bolstered the support for and confidence in the use of climate models.

Climate models may be portrayed as alarmist or extreme by some skeptics, but they actually tend to be more conservative with their predictions since a net positive feedback—that is, an overloading of the system with greenhouse gases—skews toward a stronger climate response. One example is the recorded sea level rise from satellite data collected from 1993 to 2008. CMIP3 models predicted just 1.9 millimeters of sea level rise while the data collected in the following years showed 3.4 millimeters of sea level rise. CMIP models also underestimated sea ice melt-out rates. Between 2007 and 2009, the amount of sea ice that melted was 40 percent greater than the average predicted by CMIP4 models.

One aspect of the new CMIP6 models that will work toward addressing underestimations are their higher climate sensitivity in comparison to past assessments. Climate sensitivity is the amount of long-term warming expected after a doubling of carbon emissions that greatly affects the model’s predictions. To eliminate more of the uncertainty from past models, scientists and mathematicians are constantly evolving climate models to improve their accuracy. 

Though it is undeniable that models innately have uncertainties and the Earth encompasses a collection of complex systems, climate models have proven to be reliable predictors of climate change trends. Whether or not these most recent models will continue in that vein can only be determined over time, but climate action is not something that can wait any longer.

EarthTalk® is produced by Roddy Scheer & Doug Moss for the 501(c)3 nonprofit EarthTalk. See more at To donate, visit https// Send questions to: